Re: groff maintainership, release, and blockers

[Ingo Schwarze]

Hi Bertrand,

Bertrand Garrigues via wrote on Fri, Aug 26, 2022 at 11:53:07PM +0200:
> On Fri, Aug 26 2022 at 02:04:57 PM, G. Branden Robinson wrote:
>> At 2022-08-26T13:51:25+0200, Ingo Schwarze wrote:

>>> In particular, i'm firmly convinced that issuing an RC while even one
>>> single blocker issue is unresolved is a blatant contradiction.  Before
>>> an RC, all blockers must either be resolved or explicitly
>>> re-classified as "not release critical" and re-scheduled for the
>>> subsequent release.

> I understand your point Ingo, however the rc1 tag is almost 2 years old,

Yes, no argument about the RC1 tag going wrong and being useless now.

> so I feel we need to make a new tag now, and from this tag decide
> which bugs must absolutely be fixed.

I'm not convinced deciding which issues shall be fixed and which postponed
needs a tag, nor that calling such a tag an "RC" is easy to understand,
but i think Branden is right that's only a terminolgical argument
after all.

> I won't release any official 1.23.0 if you consider there is a blocker
> or that the mandoc is not in a good shape.

Actually, i deem the state of mandoc irrelevant in this context.  It is
perfectly fine to release groff while mandoc is in an unstable state.
(As a matter of fact, mandoc is not close to release-ready right now:
it has various open bugs and some unfinished work, but that is off-topic
here.)

> For sure there will be some bug fixes after rc2 and we'll have an rc3,
> so it's perhaps not exactly a "Release Candidate", it's a kind of
> "intermediate tag" or an "alpha release", but I'll still name it rc2,
> for the sake of simplicity.

Fair enough, file name consistency is a plus.
It is desirable for the announcement be clear about the intended
purpose of RC2, though.

> Could you please detail here what is your list of blockers that you
> think must be absolutely fixed before the official 1.23.0 release?

I cannot give a final list off the top of my head because i am aware
of several candiadates of regressions that require analysis.

Are you planning to move towards an RC during the next two to three
weeks?  In that case, i should probably priotize triage over fixing
which i usually avoid (if possible) because it increases the total
working time consumed.  But it would speed up drafting a list of
potential issues that might be worth fixing before release.


I recommend investigating this one before RC2, if feasible:

#62918 Wrong GhostScript version reported during build

There remains a regression in man(7) .EX/.EE which i will report ASAP.

Even though this is documentation only, i recommend resolving it before
RC3 because switching back and forth in terminology in three consecutive
releases could be regarded as awkward:

#62816 rename the \& escape...again

I recommend postponing these until after release:

#62774 [mdoc] warn if any of `Dd`, `Dt`, `Os` not called
#62926 [mdoc] align styling of titles and man page cross references with man(7)
#62933 [man] produce hyperlinks in PDF output

Yours,
  Ingo